Transfer Learning in Python? - with practical example

Transfer Learning in Python is a technique that allows us to leverage pre-trained models to solve new tasks with less data and computational resources. This is especially useful when working with limited datasets or when training deep learning models from scratch is not feasible. **Example 1: Fine-tuning a Pre-trained Image Classification Model** Step 1: Load a pre-trained model (e.g., ResNet) and freeze its layers to prevent them from being updated during training. Step 2: Replace the final classification layer with a new one that matches the number of classes in your new dataset. Step 3: Unfreeze the top layers of the model and train the entire network on the new dataset. 
import tensorflow as tf
from tensorflow.keras.applications import ResNet50
from tensorflow.keras.layers import Dense
from tensorflow.keras.models import Model

base_model = ResNet50(weights='imagenet', include_top=False)
for layer in base_model.layers:
    layer.trainable = False

x = base_model.output
x = tf.keras.layers.GlobalAveragePooling2D()(x)
x = Dense(1024, activation='relu')(x)
predictions = Dense(num_classes, activation='softmax')(x)

model = Model(inputs=base_model.input, outputs=predictions)
model.compile(optimizer='adam', loss='categorical_crossentropy', metrics=['accuracy'])

model.fit(train_data, train_labels, epochs=10, validation_data=(val_data, val_labels))

**Example 2: Feature Extraction using a Pre-trained Model** Step 1: Load a pre-trained model (e.g., VGG16) without the top classification layers. Step 2: Extract features from the pre-trained model for your dataset. Step 3: Train a new classifier (e.g., SVM) on the extracted features. 
import numpy as np
from tensorflow.keras.applications import VGG16
from tensorflow.keras.applications.vgg16 import preprocess_input
from tensorflow.keras.preprocessing import image
from sklearn.svm import SVC

base_model = VGG16(weights='imagenet', include_top=False)
feature_extractor = Model(inputs=base_model.input, outputs=base_model.get_layer('block5_pool').output)

X_train_features = []
for img_path in X_train:
    img = image.load_img(img_path, target_size=(224, 224))
    x = image.img_to_array(img)
    x = np.expand_dims(x, axis=0)
    x = preprocess_input(x)
    features = feature_extractor.predict(x)
    X_train_features.append(features)

svm = SVC()
svm.fit(X_train_features, y_train)

By following these steps, you can effectively utilize transfer learning to improve the performance of your models, even with limited data. Transfer learning is a powerful tool in the field of deep learning that can help you achieve better results in a more efficient manner.

Comments

Post a Comment

Popular posts from this blog

What are the different types of optimization algorithms used in deep learning?

In the first example, we are going to discuss the Stochastic Gradient Descent (SGD) optimization algorithm in deep learning using Python. # Import necessary libraries import numpy as np from sklearn.datasets import make_classification from sklearn.linear_model import SGDClassifier from sklearn.model_selection import train_test_split from sklearn.preprocessing import StandardScaler # Create a synthetic dataset X, y = make_classification(n_samples=1000, n_features=20, n_classes=2, random_state=42) # Split the dataset into training and testing sets X_train, X_test, y_train, y_test = train_test_split(X, y, test_size=0.2, random_state=42) # Standardize features by removing the mean and scaling to unit variance scaler = StandardScaler() X_train = scaler.fit_transform(X_train) X_test = scaler.transform(X_test) # Initialize the SGDClassifier with appropriate parameters sgd_clf = SGDClassifier(loss='log', max_iter=1000, learning_rate='optimal') # Fit the model on the train...
Read more

What are the different evaluation metrics used in machine learning?

In the first example, we are going to demonstrate how to use accuracy as an evaluation metric in machine learning using Python. Here is the code: from sklearn.metrics import accuracy_score y_true = [0, 1, 1, 0, 1] y_pred = [0, 1, 0, 0, 1] accuracy = accuracy_score(y_true, y_pred) print("Accuracy:", accuracy) Step 1: Import the accuracy_score function from sklearn.metrics. Step 2: Define two lists, y_true and y_pred, containing the true and predicted values. Step 3: Calculate the accuracy score by calling the accuracy_score function with y_true and y_pred as arguments. Step 4: Print out the accuracy score. In the second example, we are going to demonstrate how to use precision and recall as evaluation metrics in machine learning using Python. Here is the code: from sklearn.metrics import precision_score, recall_score y_true = [0, 1, 1, 0, 1] y_pred = [0, 1, 0, 0, 1] precision = precision_score(y_true, y_pred) recall = recall_score(y_true, y_pred) print("Precis...
Read more

What is the difference between a module and a package in Python?

In the first example, we are going to explore the concept of a module in Python. A module is a file containing Python code that can define functions, classes, and variables. It allows you to organize your code into reusable components. Here is an example code snippet demonstrating the use of a module in Python: # Create a module named mymodule.py # Define a function inside the module def greet(name): return "Hello, " + name In the code above, we have created a module named mymodule.py that defines a function called greet . This function takes a name as an argument and returns a greeting message. To use this module in another Python script, we can import it using the import keyword. import mymodule message = mymodule.greet("Alice") print(message) When we run the above code, it will import the mymodule module and use the greet function to print a greeting message for the name "Alice". In the second example, we are going to discuss the ...
Read more